ABSTRACT Approximately 100,000 women per year are diagnosed with atypical hyperplasia (AH) of the breast, a benign breast lesion that is associated with a fourfold increase in risk of subsequent breast cancer. Usage of tamoxifen as cancer prevention therapy following AH diagnosis has been linked to a greater than 50% reduction in subsequent breast cancer incidence. However, uptake of tamoxifen and similar prevention therapies remains low, due in part to a lack of methods to accurately determine which women are at greatest risk for breast cancer (BC), and which women will have a beneficial response to prevention therapy. We propose a highly translational project to address these critical barriers. We will accomplish this by identifying molecular features in benign breast tissue that will 1) improve individualized BC risk prediction for women with AH, 2) serve as biomarkers associated with beneficial response to prevention therapy, and 3) permit assessment of individualized response to prevention therapy, even after a short period of treatment. In Specific Aim 1, we will develop a breast cancer risk prediction model for women with AH that incorporates a NanoString-based gene expression assay in combination with clinical and histological variables; in Specific Aim 2, we will utilize a newly developed benign breast tissue cohort to identify biomarkers that predict beneficial response to tamoxifen; in Specific Aim 3, we will conduct a randomized clinical trial of tamoxifen versus a novel tamoxifen metabolite with less toxicity. This trial in women with AH will determine if beneficial biomarker responses to tamoxifen are observed after only four weeks of treatment. Together, our proposed studies will develop biomarkers that will have immediate relevance to women with AH, and implementation of these approaches in the clinic will have a powerful and sustained impact on the field of BC prevention.